Investigate brain volume changes upon a vasodilatory hypercapnic stimulus as a potential alternative to obtain cerebral tissue reactivity information.Measurements of cerebrovascular reactivity (CVR) with hypercapnic stimuli are commonly used for the assessment of cerebrovascular reserve capacity in patients with cerebrovascular disease^1,2. Most popular methods for obtaining spatial and temporal CVR information are BOLD and ASL MRI. Generally these methods perform adequately in healthy subjects, but clear interpretation of BOLD and ASL signals can become ambiguous in cases of severe hemodynamic impairment^3,4,5. BOLD signal changes originate from a complex interaction of CBF, CBV, and CMRO2 changes and are dependent on the baseline physiological state including baseline Yv, OEF, and CBF⁶. For ASL MRI, quantification of CBF CVR responses can be problematic in areas with severely prolonged bolus arrival times and under changed conditions of arterial blood oxygenation such as hypoxia. The aim of this study was to investigate whether brain volume changes can be observed upon a hypercapnic stimulus using relatively standard 3D T1-weighted scans (MP-RAGE) and segmentation software. This approach could yield novel insights on cerebral reactivity under hypercapnia and be a potential alternative for obtaining cerebral reactivity information in cases where BOLD or ASL CVR interpretation can be ambiguous.Healthy volunteers (n=8; age=29+/-9 mean +-/s.d.) were scanned at 3T (Philips) with an 8 channel SENSE-head coil. The following MP-RAGE acquisition parameters were used: sagittal 3D IR-TFE (SPGR), voxel size=1x1x1mm³, SENSE 2, FOV=240x240x180mm³, matrix size=240x240, TR=8ms, TE=3.2ms, TI=950ms, flip angle: 10°, BW=191 Hz, shot-interval=2100ms, acquisition time=3min11s. Two MP-RAGE scans were acquired; during normoxic normocapnia (baseline scan) and during normoxic hypercapnia. The normoxic hypercapnic challenge consisted of a boxcar stimulus in which EtCO2 was targeted by 10 mmHg above resting baseline EtCO2 over a period of 4min30s, where after 1 min the MP-RAGE scan was started. Hypercapnic gasses were administered using a computer-controlled rebreathing method (RespirAct^TM, Thornhill Research Inc.). Brain volume changes were assessed using Freesurfer^7,8,9. Ventricular volume changes were assessed with VIENA¹⁰ using manually delineated masks of the lateral ventricles. Freesurfer segmentation results revealed subcortical deep gray matter volume increases for all subjects (p=0.0025, Figure 1A). These changes were paralleled by a significant decrease in lateral ventricles volume for all subjects (p=9e-4, Figure 1B). Other regions assessed by Freesurfer did not show a significant difference between normoxia and hypercapnia at a level of p=0.05. An example of manually delineated lateral ventricles are shown in Figure 2 (Figure 1A; mask overlaid on a slice of an acquired T1-weighted MP-RAGE scan, Figure 1B 3D; rendering of the same mask). A deep gray matter volume change was found during hypercapnia. This may be caused by the high perfusion of the deep gray matter combined with the robustness of the volume measurements in these deep gray matter regions. An explanation of the ventricular volume decrease is the possible compression of the ventricles as an passive reaction of the deep gray matter volume increase. Accompanying assessment of ventricular CSF outflow changes during hypercapnia can be of interest here. The observed change in brain volume may provide information on the flexibility of the brain to change the water fraction from one compartment to another. In this respect the ability to change in volume between brain tissue/CSF compartments in short periods of time can be anticipated to be an aspect of a healthier brain. Future research is warranted to determine the exact cause of the observed volume changes and the importance of the different compartments outside the CSF (cerebral blood volume, interstitial water and cellular water). Ongoing work is aimed at increasing spatial resolution and tissue contrast within an acceptable scan time for hypercapnic exposure (both at 3T and 7T). Finally, this novel concept of volume reactivity may be a new biomarker of assessing brain tissue flexibility in healthy and disease.